Towards Safe Maneuvering of Double-Ackermann-Steering Robots with a Soft Actor-Critic Framework
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911208318173184 |
|---|---|
| author | Deflesselle, Kohio Daniel, Mélodie Magassouba, Aly Aranda, Miguel Ly, Olivier |
| author_facet | Deflesselle, Kohio Daniel, Mélodie Magassouba, Aly Aranda, Miguel Ly, Olivier |
| contents | We present a deep reinforcement learning framework based on Soft Actor-Critic (SAC) for safe and precise maneuvering of double-Ackermann-steering mobile robots (DASMRs). Unlike holonomic or simpler non-holonomic robots such as differential-drive robots, DASMRs face strong kinematic constraints that make classical planners brittle in cluttered environments. Our framework leverages the Hindsight Experience Replay (HER) and the CrossQ overlay to encourage maneuvering efficiency while avoiding obstacles. Simulation results with a heavy four-wheel-steering rover show that the learned policy can robustly reach up to 97% of target positions while avoiding obstacles. Our framework does not rely on handcrafted trajectories or expert demonstrations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_10332 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Safe Maneuvering of Double-Ackermann-Steering Robots with a Soft Actor-Critic Framework Deflesselle, Kohio Daniel, Mélodie Magassouba, Aly Aranda, Miguel Ly, Olivier Robotics Artificial Intelligence We present a deep reinforcement learning framework based on Soft Actor-Critic (SAC) for safe and precise maneuvering of double-Ackermann-steering mobile robots (DASMRs). Unlike holonomic or simpler non-holonomic robots such as differential-drive robots, DASMRs face strong kinematic constraints that make classical planners brittle in cluttered environments. Our framework leverages the Hindsight Experience Replay (HER) and the CrossQ overlay to encourage maneuvering efficiency while avoiding obstacles. Simulation results with a heavy four-wheel-steering rover show that the learned policy can robustly reach up to 97% of target positions while avoiding obstacles. Our framework does not rely on handcrafted trajectories or expert demonstrations. |
| title | Towards Safe Maneuvering of Double-Ackermann-Steering Robots with a Soft Actor-Critic Framework |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2510.10332 |